Natalia Ivanova ¹, Radu Dobrin ¹, Rong Lu ¹, Iulia Kotenko ¹, John Levorse ¹, Christina DeCoste ¹, Xenia Schafer ¹, Yi Lun ¹, and Ihor R. Lemischka ¹

¹ Department of Molecular Biology, Princeton University, Princeton, New Jersey 08544, USA

Correspondence and requests for materials should be addressed to:
N.I.:  Email: nivanova@molbio.princeton.edu   or
I.R.L.:  Email: ilemischka@molbio.princeton.edu

Abstract:

We present an integrated approach to identify genetic mechanisms that control self-renewal in mouse embryonic stem cells. We use short hairpin RNA (shRNA) loss-of-function techniques to downregulate a set of gene products whose expression patterns suggest self-renewal regulatory functions. We focus on transcriptional regulators and identify seven genes for which shRNA-mediated depletion negatively affects self-renewal, including four genes with previously unrecognized roles in self-renewal. Perturbations of these gene products are combined with dynamic, global analyses of gene expression. Our studies suggest specific biological roles for these molecules and reveal the complexity of cell fate regulation in embryonic stem cells.

Mouse embryonic stem (ES) cells can differentiate into cells of all three germ layers, and self-renew extensively in culture [1]. The transcription factors Oct4 (also known as Pou5f1) and Nanog, as well as the LIF–gp130–Stat3, BMP–TGF-beta–Smad, MAPK–ERK and possibly the WNT signalling pathways, all have important roles [2, 3, 4, 5, 6, 7, 8, 9]. Cell fate choices in ES cells are clearly regulated by a complex orchestration of multiple pathways. Because this complexity is poorly understood, it has been difficult to produce specific lineages from ES cells and to re-programme somatic cell genomes after nuclear transplantation [10, 11]. Here we dissect self-renewal using an integrated functional genomics strategy. We identify seven self-renewal regulators, suggest specific functional roles and provide a framework to analyse cell fate regulatory networks in ES cells.

Identifying gene products required for self-renewal

We reasoned that like Nanog and Oct4, other genes required for self-renewal are downregulated upon induction of differentiation. Microarray time-course analyses of ES cells (six time points at 1-day intervals) undergoing retinoic-acid-induced [12] differentiation identified 901 rapidly downregulated genes (Supplementary Fig. 1). Sixty-five genes encoding transcription factors/DNA-binding proteins and unassigned expressed sequence tags (ESTs), as well as five additional retinoic-acid-insensitive genes identified in our previous study [13], were chosen and downregulated using shRNA delivered by lentiviral vectors (Fig. 1a and Supplementary Fig. 2). The shRNA is expressed from an H1 promoter, while a constitutive ubiquitin C promoter drives hygromycin–green fluorescent protein (GFP) expression. More than 90% ES cell transduction is routinely achieved without drug selection. For a complete description see Supplementary Information.

Figure 1: Identification of self-renewal regulators in ES cells.

Figure 1 : Identification of self-renewal regulators in ES cells.

a, Lentiviral vectors expressing both a shRNA and a hygromycin–EGFP fusion protein were constructed for each of the selected genes. FLAP, nucleotide segment that improves transduction efficiency; dLTR, deleted long-terminal repeat; WRE, woodchuck hepatitis virus post-transcriptional regulatory element.

b, A competition strategy was designed to identify gene products whose depletion compromises ES cell self-renewal. GFP+ (shRNA-expressing) cells were mixed in a 4-to-1 ratio with untransduced (GFP-) cells, and cultured in the presence of LIF. With each passage the ratio of GFP+/total cells was measured by flow cytometry.

c, Depletion of Nanog and Oct4 as well as eight additional gene products results in impaired self-renewal. The proportions of GFP+ cells transduced with the empty H1P vector do not change (representative experiment shown). In all cases Real-Time PCR (RT–PCR) confirms the reduction of mRNA levels (residual amounts relative to wild-type levels are shown as percentages above each time course).

Progressive decreases in GFP+/GFP- ratios were observed upon downregulation of 10 out of 70 genes (Fig. 1c). These are: transcription factors Nanog, Oct4, Sox2, Tbx3, Esrrb[2, 4, 15, 16, 17]; a cofactor of the Akt1 kinase Tcl1 (ref. 18); an uncharacterized ES cell-specific gene Dppa4 (ref. 11); and unassigned ESTs Mm.343880, Mm.276044 and Mm.219358. Target messenger RNA levels were significantly decreased by all ten shRNAs. For detailed information see Supplementary Table 1.

Downregulation of some genes could retard the cell cycle without inducing differentiation. Accordingly, we evaluated cell morphology and alkaline phosphatase activity as indicators for undifferentiated cells. Downregulation of Nanog, Oct4, Sox2, Tbx3, Esrrb, Tcl1, Dppa4 and Mm.343880 resulted in morphological changes and loss of alkaline phosphatase activity (Supplementary Fig. 4). No colonies were obtained after downregulation of Mm.276044 and Mm.219358, suggesting roles for these genes in the control of cell cycle or viability. In other cases, small alkaline-phosphatase-positive colonies were obtained and were not analysed further (data not shown). Morphological evaluation and alkaline phosphatase assays were performed for all 60 shRNA vectors that did not cause a change in GFP+/GFP- ratios. In no case was the morphology or alkaline phosphatase activity changed, indicating that the competition assay is a robust way to detect ES cell differentiation.

Complementation rescue of compromised self-renewal activity

To rule out nonspecific effects of shRNAs we designed three additional shRNAs for each gene (Supplementary Table 1). At least one additional shRNA was effective for Nanog, Oct4, Sox2, Esrrb and Tcl1, showing that off-target effects are not responsible for the observed phenotypes. We also devised a genetic complementation (rescue) strategy where the original shRNA vectors were modified to include tetracycline-inducible versions of the shRNA-targeted gene products (Fig. 2a). To facilitate this rescue strategy the original shRNA targets were chosen to be in the 3' untranslated region (with the exception of Oct4 and Tcl1). The shRNA-immune versions contain only the protein-coding sequences. Transactivator (rtTA)-expressing ES cells were transduced [19]. Self-renewal in transduced cells is doxycycline-dependent, as shown in Fig. 2b for Nanog. Nanog rescue-vector-transduced GFP+ cells (NanogR) were cloned in the presence of doxycycline and compared to clones transduced with empty vector (ControlR). Endogenous Nanog transcript was permanently downregulated, whereas vector-derived transcript was induced in the presence of doxycycline. The latter was undetectable 2 days after doxycycline removal. By morphological and alkaline phosphatase criteria, NanogR clones were undifferentiated in the presence of doxycycline. Upon doxycycline removal NanogR clones underwent morphological changes and lost alkaline phosphatase activity. Morphology and alkaline phosphatase activity of ControlR clones were unaffected by doxycycline (Fig. 2c). Robust doxycycline-dependent self-renewal was observed in Sox2, Esrrb, Tbx3 and Tcl1 rescue experiments. In the case of Oct4, we were not able to fully restore endogenous Oct4 levels. Oct4R clones grow slowly but retain ES cell morphology, alkaline phosphatase activity and express high levels of Nanog. However, further analysis showed that several markers of trophectoderm stem (TS) cells, such as Hand1, Cdx2, Eomes and Mash2 (referenced in Supplementary Table 2), were upregulated in the presence of doxycycline (Supplementary Fig. 6). These cells probably represent an intermediate stage between ES and TS cells. After removal of doxycycline, Oct4R cells become negative for alkaline phosphatase, further upregulate TS markers and acquire characteristic trophoblast morphology. In the case of Dppa4, all rescue clones showed strong downregulation of endogenous mRNA but yielded a significant number of undifferentiated colonies (data not shown). Therefore, Dppa4 may not be absolutely required for self-renewal. No rescue clones were obtained for Mm.343880, suggesting that the morphological changes observed in the initial shRNA experiments were potentially due to the off-target effects or technical complications associated with vector packaging (Supplementary Fig. 5). This gene was excluded from most subsequent analyses. To confirm further the differentiated state of cells maintained without doxycycline we measured the expression levels of pluripotency markers Nanog and Oct4. After 8 days without doxycycline both markers were significantly reduced in NanogR, Sox2R, EsrrbR, Tbx3R and Tcl1R cells (Fig. 2e).

Figure 2: Genetic complementation to rescue shRNA-induced self-renewal defects.

Figure 2 : Genetic complementation to rescue shRNA-induced self-renewal defects.

a, The shRNA vectors were modified to produce tetracycline-inducible versions of endogenous targets. EGFP is coupled to the overexpressed gene via an internal ribosome entry sequence (IRES) (Nanog rescue vector is shown as an example).

b, This approach was tested using Nanog. The rescue vectors were introduced into rtTA transactivator expressing ES cells in the presence of doxycycline (Dox); individual clones were isolated and expanded. Three clones (labelled 1–3) were chosen for further analysis. Control (ControlR) clones transduced with empty rescue vector (H1P-pTRE-IRES-EGFP) are also shown. Levels of endogenous Nanog (Nanog RNA) and vector derived Nanog (Nanog Lenti) were measured by northern blotting at 2-day intervals for 6 days (d0–d6) after withdrawal of doxycycline. In all three NanogR clones endogenous Nanog mRNA was permanently decreased. Vector-derived transcripts were expressed in the presence of doxycycline and rapidly reduced after doxycycline withdrawal.

c, Self-renewal was rescued in the case of Nanog and for six other gene products.

d, Levels of endogenous (RNA) and vector-derived (Lenti) transcripts were measured for three independent clones with and without doxycycline for the six additional genes. In all cases, the levels of endogenous transcript were decreased, and expression of the vector-derived transcripts was dependent on doxycycline.

e, Expression of self-renewal markers Nanog and Oct4 was reduced in the rescue cells maintained without doxycycline for 8 days. Expression level before the removal of doxycycline is shown as d0. Expression in d8 ControlR cells maintained with doxycycline is set as 100%. Each bar represents the average expression in three independent rescue clones corresponding to each gene; standard deviations are shown as error bars.

Self-renewal regulators suppress differentiation

Loss of individual self-renewal regulators may initiate distinct differentiation programmes. We measured expression of early trophectodermal, mesodermal, ectodermal and endodermal markers in the rescue clones during an 8-day period after doxycycline removal. In addition, the expression of markers for more mature cell types that originate from each germ layer was measured at day 8. In total, a set of 35 markers was analysed (Supplementary Table 2). All polymerase chain reaction with reverse transcription (RT–PCR) analyses were performed using two independent rescue clones. Observed marker patterns are shown in Supplementary Figs 6–9.

Trophectodermal markers Hand1, Cdx2, PL1, Mash2 and Ehox were induced in Oct4R, Sox2R and NanogR cells during the 8-day period after doxycycline removal. These markers were not induced when NanogR and Sox2R cells were maintained with doxycycline. The TS cell marker Eomes was upregulated only in Oct4R cells (Supplementary Fig. 6).

Goosecoid (Gsc) and brachyury (T), markers for primitive streak and early mesoderm, were transiently expressed in NanogR, Sox2R, EsrrbR and Tbx3R cells. In addition, Mixl1, expressed in cells of the organizer region that will become mesoderm, was induced in NanogR and Sox2R cells. After the expression of early mesodermal markers, expression of genes specific to mesoderm-derived cell lineages was detected. These are: Cd34 and beta-globin (extraembryonic mesoderm); Bmp5, Shh and Tbx5 (dorsal mesoderm); Gata5 and Isl1 (cardiac mesoderm and definitive endoderm); Meox1 and Cart1 (lateral mesoderm). The induction of these markers was not observed when cells were maintained with doxycycline (Supplementary Fig. 7).

Expression of the primitive ectoderm marker Fgf5 was transiently induced in NanogR, Sox2R, EsrrbR, Tbx3R and Tcl1R cells, and was followed by upregulation of the ectodermal marker Cxcl12. Mash1 and Pax3 but not Sox1 were also induced, indicating that these cells differentiate into neural crest derivatives. Other neural crest markers such as Ednra, Eya2, Ngn2, Phox2b and Slug were also detected in these cells. Only a subset of these markers was induced in Tcl1R cells after doxycycline removal, suggesting that additional signals are required to sustain differentiation induced by downregulation of Tcl1 (Supplementary Fig. 8).

We found that ControlR cells spontaneously upregulated endodermal markers when maintained without passaging for more than 6 days. Therefore, endodermal commitment was analysed 4 days after doxycycline removal. At this time endodermal markers (Gata4, Gata6, Foxa2, Sox17, Nr2f1, Nr2f2 and Bmp2) were induced in NanogR, Sox2R and EsrrbR cells (Supplementary Fig. 9). The markers were induced at the same time or after the induction of the organizer gene Gsc. Therefore, these cells may be differentiating into definitive endoderm from Gsc+/Foxa2+ precursors [20]. Collectively, our results demonstrate that in addition to both Oct4 and Nanog, Sox2, Esrrb, Tbx3 and Tcl1 function to suppress ES cell differentiation in vitro. None of the differentiation programmes was induced after downregulation of Dppa4. We also confirmed that removal of doxycycline does not induce apoptosis or cell cycle arrest in any of the rescue cells (Supplementary Figs 17 and 18).

To explore possible interactions between the identified gene products and known signalling pathways implicated in the control of self-renewal, we asked whether the activation status of LIF, ERK1/2, BMP and WNT signalling pathways is altered by the shRNA treatments (Supplementary Fig. 12). Stat3, Smad1/5/8 and Smad2/3 phosphorylation as well as total levels of beta-catenin were unaffected by any shRNAs. ERK1/2 was hyper-phosphorylated after downregulation of Nanog, Oct4, Sox2 and Mm.343880. The ERK1/2 pathway is activated in trophectoderm; therefore the increased phosphorylation of ERK1/2 correlates with the induction of the trophectodermal markers by Nanog, Oct4 and Sox2 shRNAs [21, 22].

Constitutive expression of self-renewal regulators

We next asked whether enforced expression of identified self-renewal regulators is sufficient to block the commitment to specific lineages. Rescue clones were used to produce embryoid bodies [23] in the presence of doxycycline. Temporal profiles of markers for neuroectoderm, mesoderm and endoderm were analysed during a 12-day period (Supplementary Fig. 10). Markers for all three lineages were induced in NanogR-, Tcl1R- and Dppa4R-derived embryoid bodies. Mesodermal commitment was affected in both Sox2R and Tbx3R cells. Marker patterns suggest that mesodermal development is blocked after the induction of Gsc in Sox2R cells (no Mixl1 and no T expression) and slightly later in Tbx3R cells (normal Mixl1 and no T expression). A complete block of both mesodermal and neuroectodermal differentiation was observed in EsrrbR cultures, where neuroectodermal differentiation was arrested after the induction of Fgf5, and mesodermal differentiation was blocked before the induction of Gsc. EsrrbR cells retained the capacity to form endoderm. In addition, expression of Sox1 was increased in Sox2R cells and probably reflects the importance of Sox2 in the neuroectodermal lineage [24]. Enforced expression of Tcl1 may also promote the expansion of neuroectodermal precursors and/or prevent their differentiation.

We also analysed the capacity of rescue cells to contribute to different tissues (brain, liver, muscle and coat hair) in vivo. ControlR cells contributed efficiently to all tissues, including the germ line, demonstrating that the rescue clone derivation procedures do not compromise pluripotency. NanogR, SoxR, Tbx3R and Tcl1R cells maintained without doxycycline for three passages did not contribute to any tissue, whereas EsrrbR cells retained limited capacity to contribute to neural crest derivatives such as coat melanocytes (see Supplementary Information for a detailed description).

shRNA-induced gene expression dynamics

We next analysed transcriptome dynamics after downregulating the expression of each of the seven genes identified in the screen. shRNA-transduced GFP+ cells were purified by fluorescence-activated cell sorting (FACS) daily during a 7-day interval and were used to interrogate Affymetrix microarrays. We identified a total of 3,109 non-redundant genes whose expression was perturbed after treatment with at least one shRNA (see Supplementary Information). Several distinct patterns of gene expression were observed. The first pattern (Fig. 3a) contains 771 genes that are up- or downregulated in response to most shRNA treatments. The second pattern (Fig. 3b) contains 474 genes perturbed in response to Nanog, Oct4 and Sox2 but not to Esrrb, Tbx3, Tcl1 or Dppa4 shRNA treatments. Recent chromatin studies have shown that Nanog, Oct4 and Sox2 often co-occupy the promoters of target genes [25]. The third pattern (Fig. 3c) contains 272 gene products that are responsive to Esrrb, Tbx3, Tcl1 and Dppa4 shRNAs, but not to Nanog, Oct4 or Sox2 shRNAs. Patterns 2 and 3 suggest the existence of at least two distinct pathways that are necessary for self-renewal. Interestingly, whereas pattern 2 contains equal numbers of up- and downregulated gene products, pattern 3 contains gene products that are preferentially upregulated by shRNA treatments. This suggests that the corresponding pathway functions to repress differentiation-promoting genes. We have identified a number of smaller gene clusters that were perturbed by single shRNA treatments or by pairs of shRNA treatments (Supplementary Fig. 13).

Figure 3: Global gene expression changes after downregulation of individual self-renewal regulators.

Figure 3 : Global gene expression changes after downregulation of individual self-renewal regulators.

Microarray analysis time courses were performed to measure gene expression changes that accompany differentiation.

a–c, Clustering revealed distinct patterns of gene expression. A number of transcriptional regulators present in the gene expression patterns that are upregulated upon shRNA-mediated depletion of self-renewal genes can induce differentiation when overexpressed in wild-type ES cells.

d, Differentiation was apparent from the morphologies of the cells as well as the loss of Nanog expression (data not shown).

e, Temporal expression profiles of identified differentiation inducers upon shRNA-mediated inactivation of individual self-renewal regulators show that the genes (listed on the right side of the panel) are upregulated in at least one, and usually more than one, shRNA time course. Expression levels of the genes were centred and scaled to one standard deviation, and are indicated by colour, from green (low expression) to red (high expression).

Identification of gene products that promote differentiation

A number of transcriptional regulators are present in the gene expression patterns defined above. Some of these are upregulated after shRNA depletion of the self-renewal regulatory genes and may act as master positive regulators of differentiation. To identify such regulators, 160 gene products for which expression profiles suggest differentiation-inducing activity were overexpressed in ES cells. In 18 cases differentiation was suggested by changes in cell morphology, loss of alkaline phosphatase activity and downregulation of Nanog (Fig. 3d and Supplementary Fig. 14). Temporal expression profiles of these 18 genes upon shRNA-mediated inactivation of self-renewal regulators (Fig. 3e) suggest that many differentiation inducers are under the control of multiple self-renewal regulators.

Nanog compensates for loss of several self-renewal regulators

Most of the differentiation regulators identified in the overexpression screen are upregulated by Nanog shRNA treatment. Nanog is a strong positive regulator of self-renewal and counteracts removal of LIF and BMP [4, 7]. We asked whether Nanog could block the differentiation induced by downregulation of the other genes identified in our studies. Nanog was overexpressed in ES cells as shown in Fig. 4a. This provides an approximately threefold increase in total Nanog levels and confers LIF independence (Supplementary Fig. 15). These cells were transduced with shRNA vectors, maintained in the presence of LIF, and evaluated using morphological criteria and alkaline phosphatase staining. Nanog overexpression restored undifferentiated morphology and alkaline phosphatase activity levels in Esrrb, Tbx3, Tcl1 and Dppa4 shRNA-transduced cells. Differentiation induced by depletion of Oct4 and Sox2 was unaffected (Fig. 4b and Supplementary Fig. 16). Therefore, the overexpression of Nanog may block differentiation into mesodermal, ectodermal and neural crest cell lineages. Nanog may block differentiation by simply upregulating the endogenous targeted gene to levels that are insensitive to shRNA depletion. However, in all cases the endogenous genes were similarly depleted both in Nanog overexpressing (NanogIP) and control (ControlIP) cells (Fig. 4c). Instead, the expression levels of other self-renewal regulators that are normally downregulated by shRNA treatments were fully or partially restored. Specifically, Oct4 and Sox2 were downregulated by Esrrb shRNA in ControlIP cells, but were maintained in NanogIP cells. Nanog overexpression restored the levels of Oct4, Sox2 and Esrrb in Tbx3 shRNA-treated cells and the levels of Esrrb and Tbx3 in Tcl1 shRNA-treated cells. Overexpression of Nanog was sufficient to prevent the induction of several differentiation inducers. Cited1, Fos and Irx3 genes are shown in Fig. 4c as examples of such regulation.

Figure 4: Nanog rescue of phenotypes caused by shRNA directed against other self-renewal regulators.

Figure 4 : Nanog rescue of phenotypes caused by shRNA directed against other self-renewal regulators.

a, ES cells were transduced with a Nanog overexpressing vector. Nanog overexpressing (NanogIP, N) and control (ControlIP, C) cells were challenged with the entire set of shRNA vectors.

b, The cells were propagated in the presence of LIF, and analysed by morphology and alkaline phosphatase staining.

c, Enforced Nanog activity does not simply upregulate the targeted endogenous genes to levels insensitive to shRNA inhibition but may also compensate for the loss of function of the targeted gene by maintaining the expression of other self-renewal regulators (blue) and by preventing the induction of key differentiation inducers (purple) that normally occurs upon shRNA-mediated downregulation of Esrrb, Tbx3 or Tcl1. Gene expression levels were measured by microarrays.

Using an integrated functional genomics approach we have demonstrated that Esrrb, Tbx3 and Tcl1, as well as previously identified Nanog, Oct4 and Sox2, are required for efficient self-renewal of ES cells in vitro. Downregulation of each gene induces differentiation of ES cells along specific lineages. The data obtained from our studies are summarized in Fig. 5.

Figure 5: Provisional model of cell fate regulatory interactions in ES cells in vitro.

Figure 5 : Provisional model of cell fate regulatory interactions in ES cells in vitro.

Esrrb, Tbx3 and Tcl1, as well as previously identified Nanog, Oct4 and Sox2, are required for efficient self-renewal of ES cells in vitro. Oct4 is required to prevent trophectodermal differentiation, Nanog and Sox2 appear to be global regulators that repress multiple differentiation programmes, whereas Esrrb, Tbx3 and Tcl1 are necessary to block the differentiation into epiblast-derived lineages. Self-renewal regulators are integrated into an interconnected transcriptional network and control the expression of downstream target genes through distinct molecular pathways.

According to current models Oct4 is required to prevent trophectodermal differentiation of ES cells3, whereas Nanog is required to block differentiation to the endodermal lineage [5]. Whereas Oct4 downregulation exclusively induces the set of trophectodermal genes, we found that, in addition to endoderm, Nanog downregulation induces the expression of markers for trophectoderm and epiblast-derived lineages, namely mesoderm, ectoderm and neural crest cells. Therefore, Nanog seems to be a global regulator that represses multiple differentiation programmes. Sox2 functions to repress the development of trophectoderm and epiblast-derived lineages, whereas Esrrb and Tbx3 are necessary to block the differentiation into mesoderm, ectoderm and neural crest cells but are not required to repress trophectoderm differentiation. The function of Tcl1 seems to be even more restricted as it appears to repress only a subset of neural crest genes. In addition, enforced expression of Sox2 and Tbx3 is sufficient to repress the commitment to mesodermal lineage and Esrrb is capable of blocking both mesodermal and neuroectodermal commitment in embryoid body conditions.

We identify several patterns of transcriptional regulation. These patterns suggest the existence of two global pathways required for self-renewal and will be useful in efforts to further elucidate epistatic and other relationships. A number of genes that are upregulated after shRNA depletion of the self-renewal regulatory genes can act as positive regulators of differentiation programmes. Downregulation of Nanog, Sox2, Esrrb, Tbx3 or Tcl1 leads to the immediate induction of Otx2, Pitx2, Sox18 and probably additional genes. All three genes are expressed in the epiblast; Otx2 and Pitx2 are important for mesodermal and neuroectodermal development in vivo [26, 27, 28, 29]. Otx2 seems to be a direct transcriptional target of both Nanog and Oct4, whereas Pitx2 may be directly regulated by Nanog and Sox2, as suggested by recent chromatin-precipitation studies [25, 30]. The activation of early-induced differentiation regulators is followed by the induction of the next wave of differentiation genes. Combinatorial action of these later-acting genes specifies the precise developmental outcome, such as mesoderm, ectoderm or neural-crest-like cells observed in our experimental settings. Notably, some of the late-induced genes seem to be direct transcriptional targets of self-renewal regulators. For instance, Nanog and Oct4 are directly bound to the promoter regions of Snai1, Klf6 and Klf7 genes [30].

Nanog can substitute for Esrrb, Tbx3 and Tcl1 when overexpressed, probably through compensatory changes in the levels of other self-renewal regulators. This suggests that the self-renewal regulators are integrated into an interconnected transcriptional network and that the loss of function of one regulator can, in some cases, be compensated by adjusting the expression levels of other network components. Such a network would explain why knockouts of Esrrb, Tbx3 and Tcl1 show phenotypes at fairly advanced embryonic stages [17, 31, 32, 33, 34]. It will also be interesting to check whether Sox2, Esrrb and Tbx3 can rescue differentiation into the epiblast-derived lineages induced by Nanog shRNA treatment. The function of Dppa4 in ES cells remains unknown. None of the differentiation programmes analysed was induced in Dppa4R cells after removal of doxycycline, and the enforced expression of Dppa4 did not affect differentiation in the embryoid body assay.

The screening strategy developed in this study can be extended to genome-scale shRNA libraries to identify additional gene products that have important roles in ES cell self-renewal [35, 36]. It can also be applied to other systems, such as haematopoietic and neural stem cells, where developmental read-outs can be readily measured.

Regulatory similarities in mouse and human ES cells are unclear. Oct4 and Nanog are essential; however, the LIF–gp130–Stat3 pathway appears not to be required in human ES cells [37]. It will be important to evaluate the panel of genes described herein as candidate regulators in the human ES system.

Methods:

A detailed description of the methods used in these studies can be found in Supplementary Information.
http://www.nature.com/nature/journal/vaop/ncurrent/suppinfo/nature04915.html

This file contains Supplementary Materials, Supplementary Tables 1–4, Supplementary Figures 1–18 and additional references.
Supplementary Notes - Download PDF file (2.2MB)

Supplementary Data 1
MAS5.0 normalized microarray data
Supplementary Data 1 - Download Zip file (16.1MB)

Supplementary Data 2
MIAME description of microarray data
Supplementary Data 2 - Download Word file (55KB)

Supplementary Data 3
Description of microarray files
Supplementary Data 3 - Download Excel file (27KB)

Supplementary Data 4
Non-redundant probes used in microarray analysis
Supplementary Data 4 - Download Excel file (3.6MB)

Supplementary Data 5
Probes affected at least in one shRNA time-course
Supplementary Data 5 - Download Excel file (503KB)

Supplementary Data 6
901 probes down-regulated in RA differentiation time-course
Supplementary Data 6 - Download Excel file (732KB)

Supplementary Data 7
Information for genes used in the over-expression studies
Supplementary Data 7 - Download Excel file (43KB)

Supplementary Data 8
List of genes affected by Nanog shRNA
Supplementary Data 8 - Download Excel file (252KB)

Supplementary Data 9
List of genes affected by Oct4 shRNA
Supplementary Data 9 - Download Excel file (189KB)

Supplementary Data 10
List of genes affected by Sox2 shRNA
Supplementary Data 10 - Download Excel file (228KB)

Supplementary Data 11
List of genes affected by Esrrb shRNA
Supplementary Data 11 - Download Excel file (149KB)

Supplementary Data 12
List of genes affected by Tbx3 shRNA
Supplementary Data 12 - Download Excel file (136KB)

Supplementary Data 13
List of genes affected by Tcl1 shRNA
Supplementary Data 13 - Download Excel file (100KB)

Supplementary Data 14
List of genes affected by Mm.343880 shRNA
Supplementary Data 14 - Download Excel file (111KB)

Supplementary Data 15
List of genes affected by Dppa4 shRNA
Supplementary Data 15 - Download Excel file (74KB)

Supplementary Data 16
List of genes down-regulated during the Dppa4 shRNA time-course only.
Supplementary Data 16 - Download Excel file (20KB)

Supplementary Data 17
List of genes down-regulated during the Mm.343880 shRNA time-course only.
Supplementary Data 17 - Download Excel file (17KB)

Supplementary Data 18
List of genes up-regulated during the Mm.343880 shRNA time-course only.
Supplementary Data 18 - Download Excel file (11KB)

Supplementary Data 19
List of genes down-regulated during the Nanog and Oct4 shRNA time-courses only.
Supplementary Data 19 - Download Excel file (13KB)

Supplementary Data 20
List of genes down-regulated during the Nanog shRNA time-course only.
Supplementary Data 20 - Download Excel file (36KB)

Supplementary Data 21
List of genes up-regulated during the Nanog shRNA time-course only.
Supplementary Data 21 - Download Excel file (25KB)

Supplementary Data 22
List of genes down-regulated during the Oct4 shRNA time-course only.
Supplementary Data 22 - Download Excel file (21KB)

Supplementary Data 23
List of genes up-regulated during the Oct4 shRNA time-course only.
Supplementary Data 23 - Download Excel file (20KB)

Supplementary Data 24
List of genes up-regulated during the Oct4 and Sox2 shRNA time-courses only.
Supplementary Data 24 - Download Excel file (14KB)

Supplementary Data 25
List of genes down-regulated in Pattern1
Supplementary Data 25 - Download Excel file (27KB)

Supplementary Data 26
List of genes up-regulated in Pattern1
Supplementary Data 26 - Download Excel file (116KB)

Supplementary Data 27
List of genes down-regulated in Pattern2
Supplementary Data 27 - Download Excel file (48KB)

Supplementary Data 28
List of genes up-regulated in Pattern2
Supplementary Data 28 - Download Excel file (48KB)

Supplementary Data 29
List of genes down-regulated in Pattern3
Supplementary Data 29 - Download Excel file (15KB)

Supplementary Data 30
List of genes up-regulated in Pattern3
Supplementary Data 30 - Download Excel file (49KB)

Supplementary Data 31
List of genes down-regulated during the Sox2 shRNA time-course only
Supplementary Data 31 - Download Excel file (16KB)

Supplementary Data 32
List of genes up-regulated during the Sox2 shRNA time-course only
Supplementary Data 32 - Download Excel file (19KB)

Supplementary Data 33
List of genes down-regulated during the Tbx3 shRNA time-course only
Supplementary Data 33 - Download Excel file (18KB)

Retinoic-acid-induced differentiation, competition assays, western blot analyses, microarray analyses of shRNA-induced differentiation and rescue of shRNA-induced phenotypes by Nanog overexpression were performed using mouse cell line CCE. Overexpression of shRNA-induced genes was performed using E14/T21 cells. E14/T21 cells were made by introducing EGFP under the control of the 6-kb Nanog promoter region into E14/T cells (a gift from A. Smith) using BAC transgenic technology (I.R.L. laboratory, unpublished data). CCE and E14/T21 ES cells were maintained on gelatin-coated dishes in DMEM media supplemented with 15% FBS (Hyclone), 100 mM MEM non-essential amino acids, 0.1 mM 2-mercaptoethanol, 1 mM l-glutamine (Invitrogen) and 103 U ml-1 of LIF (Chemicon).

rtTA-expressing ES cells Aniv15 (a gift from G. Daley) were used for derivation of rescue clones, and were maintained on primary mouse embryonic fibroblasts in the above-described media supplemented with doxycycline (1 microg ml-1). For differentiation assays the cells were trypsinized and plated for 30 min on standard tissue culture dishes in order to remove primary mouse embryonic fibroblasts, and floating ES cells were collected and plated on gelatin-coated dishes.

Acknowledgments

We thank A. Smith for E14/T cells and the episomal expression system, G. Daley for Aniv15 cells, and D. Baltimore for the FUGW lentiviral expression system. We are grateful to K. A. Moore and members of the Lemischka laboratory for their assistance at various stages of these studies. We also thank A. Smith for constructive criticisms. This work was supported by funding from the NIDDK. Author Contributions N.I. and I.R.L. designed the experiments. N.I., R.L., I.K., J.L., C.D., X.S. and Y.L. performed the experiments. N.I., R.D. and I.R.L. analysed the data. N.I. and I.R.L. wrote the paper.

Competing interests statement:

The authors declared no competing interests.

Supplementary information accompanies this paper.

References:

   1. Smith, A. G. Embryo-derived stem cells: of mice and men. Annu. Rev. Cell Dev. Biol. 17, 435–462 (2001)

   2. Nichols, J. et al. Formation of pluripotent stem cells in the mammalian embryo depends on the POU transcription factor Oct4. Cell 95, 379–391 (1998)

   3. Niwa, H. , Miyazaki, J. & Smith, A. G. Quantitative expression of Oct-3/4 defines differentiation, dedifferentiation or self-renewal of ES cells. Nature Genet. 24, 372–376 (2000)

   4. Chambers, I. et al. Functional expression cloning of Nanog, a pluripotency sustaining factor in embryonic stem cells. Cell 113, 643–655 (2003)

   5. Mitsui, K. et al. The homeoprotein Nanog is required for maintenance of pluripotency in mouse epiblast and ES cells. Cell 113, 631–642 (2003)

   6. Niwa, H. , Burdon, T. , Chambers, I. & Smith, A. Self-renewal of pluripotent embryonic stem cells is mediated via activation of STAT3. Genes Dev. 12, 2048–2060 (1998)

   7. Ying, Q. L. , Nichols, J. , Chambers, I. & Smith, A. BMP induction of Id proteins suppresses differentiation and sustains embryonic stem cell self-renewal in collaboration with STAT3. Cell 115, 281–292 (2003)

   8. Burdon, T. , Stracey, C. , Chambers, I. , Nichols, J. & Smith, A. Suppression of SHP-2 and ERK signalling promotes self-renewal of mouse embryonic stem cells. Dev. Biol. 210, 30–43 (1999)

   9. Sato, N. , Meijer, L. , Skaltsounis, L. , Greengard, P. & Brivanlou, A. H. Maintenance of pluripotency in human and mouse embryonic stem cells through activation of Wnt signaling by a pharmacological GSK-3-specific inhibitor. Nature Med. 10, 55–63 (2004)

  10. Loebel, D. A. , Watson, C. M. , De Young, R. A. & Tam, P. P. Lineage choice and differentiation in mouse embryos and embryonic stem cells. Dev. Biol. 264, 1–14 (2003)

  11. Bortvin, A. et al. Incomplete reactivation of Oct4-related genes in mouse embryos cloned from somatic nuclei. Development 130, 1673–1680 (2003)

  12. Tighe, A. P. & Gudas, L. J. Retinoic acid inhibits leukemia inhibitory factor signaling pathways in mouse embryonic stem cells. J. Cell. Physiol. 198, 223–229 (2004)

  13. Ivanova, N. B. et al. A stem cell molecular signature. Science 298, 601–604 (2002)

  14. Burdon, T. , Smith, A. & Savatier, P. Signalling, cell cycle and pluripotency in embryonic stem cells. Trends Cell Biol. 12, 432–438 (2002)

  15. Avilion, A. A. et al. Multipotent cell lineages in early mouse development depend on SOX2 function. Genes Dev. 17, 126–140 (2003)

  16. Bamshad, M. et al. Mutations in human TBX3 alter limb, apocrine and genital development in ulnar-mammary syndrome. Nature Genet. 16, 311–315 (1997)

  17. Mitsunaga, K. et al. Loss of PGC-specific expression of the orphan nuclear receptor ERR-beta results in reduction of germ cell number in mouse embryos. Mech. Dev. 121, 237–246 (2004)

  18. Pekarsky, Y. et al. Tcl1 enhances Akt kinase activity and mediates its nuclear translocation. Proc. Natl Acad. Sci. USA 97, 3028–3033 (2000)

  19. Ting, D. T. , Kyba, M. & Daley, G. Q. Inducible transgene expression in mouse stem cells. Methods Mol. Med. 105, 23–46 (2005)

  20. Yasunaga, M. et al. Induction and monitoring of definitive and visceral endoderm differentiation of mouse ES cells. Nature Biotechnol. 23, 1542–1550 (2005)

  21. Corson, L. B. , Yamanaka, Y. , Lai, K. M. & Rossant, J. Spatial and temporal patterns of ERK signaling during mouse embryogenesis. Development 130, 4527–4537 (2003)

  22. Saba-El-Leil, M. K. et al. An essential function of the mitogen-activated protein kinase Erk2 in mouse trophoblast development. EMBO Rep. 4, 964–968 (2003)

  23. Keller, G. Embryonic stem cell differentiation: emergence of a new era in biology and medicine. Genes Dev. 19, 1129–1155 (2005)

  24. Ferri, A. L. et al. Sox2 deficiency causes neurodegeneration and impaired neurogenesis in the adult mouse brain. Development 131, 3805–3819 (2004)

  25. Boyer, L. A. et al. Core transcriptional regulatory circuitry in human embryonic stem cells. Cell 122, 947–956 (2005)

  26. Kurokawa, D. et al. Regulation of Otx2 expression and its functions in mouse epiblast and anterior neuroectoderm. Development 131, 3307–3317 (2004)

  27. Essner, J. J. , Branford, W. W. , Zhang, J. & Yost, H. J. Mesendoderm and left-right brain, heart and gut development are differentially regulated by pitx2 isoforms. Development 127, 1081–1093 (2000)

  28. Evans, A. L. & Gage, P. J. Expression of the homeobox gene Pitx2 in neural crest is required for optic stalk and ocular anterior segment development. Hum. Mol. Genet. 14, 3347–3359 (2005)

  29. Zhang, C. , Basta, T. & Klymkowsky, M. W. SOX7 and SOX18 are essential for cardiogenesis in Xenopus. Dev. Dyn. 234, 878–891 (2005)

  30. Loh, Y. H. et al. The Oct4 and Nanog transcription network regulates pluripotency in mouse embryonic stem cells. Nature Genet. 38, 431–440 (2006)

  31. Luo, J. et al. Placental abnormalities in mouse embryos lacking the orphan nuclear receptor ERR-beta. Nature 388, 778–782 (1997)

  32. Davenport, T. G. , Jerome-Majewska, L. A. & Papaioannou, V. E. Mammary gland, limb and yolk sac defects in mice lacking Tbx3, the gene mutated in human ulnar mammary syndrome. Development 130, 2263–2273 (2003)

  33. Kang, S. M. et al. Impaired T- and B-cell development in Tcl1-deficient mice. Blood 105, 1288–1294 (2005)

  34. Narducci, M. G. et al. TCL1 participates in early embryonic development and is overexpressed in human seminomas. Proc. Natl Acad. Sci. USA 99, 11712–11717 (2002)

  35. Berns, K. et al. A large-scale RNAi screen in human cells identifies new components of the p53 pathway. Nature 428, 431–437 (2004)

  36. Paddison, P. J. et al. A resource for large-scale RNA-interference-based screens in mammals. Nature 428, 427–431 (2004)

  37. Rao, M. Conserved and divergent paths that regulate self-renewal in mouse and human embryonic stem cells. Dev. Biol. 275, 269–286 (2004)

http://www.signaling-gateway.org/update/featured/index.html

"Stem cells: self-renewal regulators revealed".

Clare Garvey, Assistant Editor, Signaling Gateway, Nature

Embryonic stem (ES) cells are unique in their ability to either self-renew or differentiate into the cell types of all three germ layers; however, the precise mechanisms underlying this self-renewal ability are unknown. A study in Nature by Lemischka and colleagues now provides insight into the gene networks required for ES cell self-renewal.

Upon differentiation, ES cells are known to downregulate the transcription factors Oct4 and Nanog, which are required for self-renewal. Microarray analysis of ES cells undergoing differentiation reveals that there are in fact over 900 genes that are downregulated, opening the possibility that many of these may also have a role in self-renewal.

The authors chose to examine the effect of RNAi-induced knockdown of 70 of these genes. RNAi-mediated silencing compromised the self-renewal capacity of ES cells in 10 of this subset. Although some of these are known self-renewal-regulating genes, others have previously uncharacterized roles. Short hairpin RNA (shRNA) constructs of Esrrb, Tbx3, Sox2, Dppa4 and Tcl1 impaired self-renewal and induced differentiation; these effects were rescued in an inducible genetic complementation strategy that restored the expression of these genes.

The authors examined the differentiation lineage adopted by cells after knockdown of individual genes, showing that Oct4 is required to prevent trophectodermal differentiation. Nanog and Sox2 were shown to act globally, repressing multiple differentiation programs, while Esrrb, Tbx3 and Tcl1 blocked differentiation into epiblast-derived lineages.

Affymetrix microarray analysis following knockdown of these genes revealed changes in the expression of over 3,000 genes. Some of these encode transcriptional regulators that may have roles as master regulators of differentiation. Indeed, several transcription factors induced ES cell differentiation when overexpressed and most of these appeared to be under the control of multiple self-renewal regulators.

As most of these differentiation factors are upregulated after Nanog shRNA knockdown, the authors investigated if Nanog expression could restore cells to an undifferentiated state following knockdown of the other self-renewing genes isolated in this study. Nanog expression did restore an undifferentiated morphology for Esrrb-, Tbx3-, Tcl1- and Dppa-depleted cells; however, Nanog expression could not block differentiation in Oct4- and Sox2-depleted cells.

This study reveals how self-renewal regulating genes appear to be connected in a transcriptional network that governs self-renewal and differentiation programs.

Clare Garvey, Assistant Editor
Signaling Gateway, Nature Magazine

1. Kioussis D, "Gene Regulation: Kissing Chromosomes",  Nature vol. 435, no. 7042, pp. 579-580 (June 2,  2005). http://www.nature.com/nature/journal/v435/n7042/full/435579a.html

2. Frenster JH, and Hovsepian JA, "Kissing Chromosomes and Paired Sense-Antisense RNA Synthesis".
71st Cold Spring Harbor Symposium on Quantitative Biology", Program page 62, May 31-June 5, 2006.

Links to RNA and Biological Causality:

A Brief History of  Activator RNA:

Links to Euchromatin Activator RNA Reviews:
Links to Euchromatin Activator RNA Research:
Links to Ultrastructural Probes of DNase I-Sensitive Sites:
Links to RNA as a Therapeutic Agent:
Links to Hodgkin Lymphoma Immuno-Pathology:
Links to Activated T-Lymphocyte Immunotherapy:
Links to Medical Systems Biology:
Links to Selective Gene Transcription:
Links to RNA-Induced Epigenetics:
Links to RNA-Induced Embryogenesis:
Links to RNA and Biological Causality:
Links to Reprogramming and Neoplasia:

"Ultrastructural Probes of Active DNA Sites, and the RNA Activators of DNA".

For Further Information and Feedback:
Phone:  +1 650 367 6483
E-mail: frenster@euchromatin.net